Statistical inference for the doubly stochastic self-exciting process

Simon Clinet, Yoann Potiron

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

We introduce and show the existence of a Hawkes self-exciting point process with exponentially-decreasing kernel and where parameters are time-varying. The quantity of interest is defined as the integrated parameter T −1 0 T θt dt, where θt is the time-varying parameter, and we consider the high-frequency asymptotics. To estimate it naïvely, we chop the data into several blocks, compute the maximum likelihood estimator (MLE) on each block, and take the average of the local estimates. The asymptotic bias explodes asymptotically, thus we provide a non-naïve estimator which is constructed as the naïve one when applying a first-order bias reduction to the local MLE. We show the associated central limit theorem. Monte Carlo simulations show the importance of the bias correction and that the method performs well in finite sample, whereas the empirical study discusses the implementation in practice and documents the stochastic behavior of the parameters.

Original languageEnglish
Pages (from-to)3469-3493
Number of pages25
JournalBernoulli
Volume24
Issue number4B
DOIs
Publication statusPublished - 2018 Nov 1

Fingerprint

Statistical Inference
Maximum Likelihood Estimator
Local Likelihood
Bias Reduction
Asymptotic Bias
Bias Correction
Order Reduction
Time-varying Parameters
Point Process
Central limit theorem
Estimate
Empirical Study
Time-varying
Monte Carlo Simulation
kernel
First-order
Estimator

Keywords

  • Hawkes process
  • High-frequency data
  • Integrated parameter
  • Self-exciting process
  • Stochastic
  • Time-varying parameter

ASJC Scopus subject areas

  • Statistics and Probability

Cite this

Statistical inference for the doubly stochastic self-exciting process. / Clinet, Simon; Potiron, Yoann.

In: Bernoulli, Vol. 24, No. 4B, 01.11.2018, p. 3469-3493.

Research output: Contribution to journalArticle

@article{4a22a0b0e23c411f81beba0f9b982bf7,
title = "Statistical inference for the doubly stochastic self-exciting process",
abstract = "We introduce and show the existence of a Hawkes self-exciting point process with exponentially-decreasing kernel and where parameters are time-varying. The quantity of interest is defined as the integrated parameter T −1 0 T θt ∗ dt, where θt ∗ is the time-varying parameter, and we consider the high-frequency asymptotics. To estimate it na{\"i}vely, we chop the data into several blocks, compute the maximum likelihood estimator (MLE) on each block, and take the average of the local estimates. The asymptotic bias explodes asymptotically, thus we provide a non-na{\"i}ve estimator which is constructed as the na{\"i}ve one when applying a first-order bias reduction to the local MLE. We show the associated central limit theorem. Monte Carlo simulations show the importance of the bias correction and that the method performs well in finite sample, whereas the empirical study discusses the implementation in practice and documents the stochastic behavior of the parameters.",
keywords = "Hawkes process, High-frequency data, Integrated parameter, Self-exciting process, Stochastic, Time-varying parameter",
author = "Simon Clinet and Yoann Potiron",
year = "2018",
month = "11",
day = "1",
doi = "10.3150/17-BEJ966",
language = "English",
volume = "24",
pages = "3469--3493",
journal = "Bernoulli",
issn = "1350-7265",
publisher = "International Statistical Institute",
number = "4B",

}

TY - JOUR

T1 - Statistical inference for the doubly stochastic self-exciting process

AU - Clinet, Simon

AU - Potiron, Yoann

PY - 2018/11/1

Y1 - 2018/11/1

N2 - We introduce and show the existence of a Hawkes self-exciting point process with exponentially-decreasing kernel and where parameters are time-varying. The quantity of interest is defined as the integrated parameter T −1 0 T θt ∗ dt, where θt ∗ is the time-varying parameter, and we consider the high-frequency asymptotics. To estimate it naïvely, we chop the data into several blocks, compute the maximum likelihood estimator (MLE) on each block, and take the average of the local estimates. The asymptotic bias explodes asymptotically, thus we provide a non-naïve estimator which is constructed as the naïve one when applying a first-order bias reduction to the local MLE. We show the associated central limit theorem. Monte Carlo simulations show the importance of the bias correction and that the method performs well in finite sample, whereas the empirical study discusses the implementation in practice and documents the stochastic behavior of the parameters.

AB - We introduce and show the existence of a Hawkes self-exciting point process with exponentially-decreasing kernel and where parameters are time-varying. The quantity of interest is defined as the integrated parameter T −1 0 T θt ∗ dt, where θt ∗ is the time-varying parameter, and we consider the high-frequency asymptotics. To estimate it naïvely, we chop the data into several blocks, compute the maximum likelihood estimator (MLE) on each block, and take the average of the local estimates. The asymptotic bias explodes asymptotically, thus we provide a non-naïve estimator which is constructed as the naïve one when applying a first-order bias reduction to the local MLE. We show the associated central limit theorem. Monte Carlo simulations show the importance of the bias correction and that the method performs well in finite sample, whereas the empirical study discusses the implementation in practice and documents the stochastic behavior of the parameters.

KW - Hawkes process

KW - High-frequency data

KW - Integrated parameter

KW - Self-exciting process

KW - Stochastic

KW - Time-varying parameter

UR - http://www.scopus.com/inward/record.url?scp=85046749606&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85046749606&partnerID=8YFLogxK

U2 - 10.3150/17-BEJ966

DO - 10.3150/17-BEJ966

M3 - Article

AN - SCOPUS:85046749606

VL - 24

SP - 3469

EP - 3493

JO - Bernoulli

JF - Bernoulli

SN - 1350-7265

IS - 4B

ER -